I finally finished my diploma thesis. I'm in computer science and I'm a Rubik's cube fanatic, so for my thesis I wrote a computer program that solves the cube. The thesis first introduces the cube, speedcubing and its community, some solving methods, some other solver programs and results. Then it describes Hume, the "human method evaluator", showing how to use it, some results of me using it, the implementation of the program and finally its possible future and conclusions.
Hume is not an optimal or near-optimal solver. Instead, it let's the user describe a human solving method which it then evaluates. The aim is to require a minimal description, only specifying the subgoals on the way to the fully solved cube. This way the user has very little work to do, as the program does all the dirty work. Entering cases or even turn sequences for them is not necessary. Besides the method description, the puzzle is described by the user as well - Hume is not just for the standard Rubik's cube, but for other similar puzzles, too. It's not yet as powerful as I'd like it to be, but I got some good results with it already and I intend to improve it as discussed in my thesis.
You can download the diploma thesis, the program will follow soon (I want to clean it up a bit before I show it).
Last modified: May 05 2008, 13:50:53